Fluorescence Microscopy Reagents
Primary antibodies used in this study included; rabbit monoclonal antibody anti-YAP/TAZ (D24E4; Cell Signaling Technology, Catalog No-8418), rabbit polyclonal anti-TKS5 antibody (Santa Cruz Biotechnology, Catalog No: SC-7390), and rabbit polyclonal anti-TKS5 antibody (Merck, Catalog No: 09-403). Secondary antibodies used here included; peroxidase-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, Catalog No: 115-035-003), peroxidase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Catalog No: 111-035-144), and goat anti-rabbit IgG Alexa Fluor647 (Thermo Fisher Scientific, Catalog No: A32728). F-actin was stained using Phalloidin-tetramethylrhodamine B isothiocyanate (Sigma Aldrich, Catalog No: P1951). Nuclei were stained using 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI; Sigma Aldrich-Aldrich, Catalog No- D9542).
Cultured cell lines used in this study
Cancer cell lines purchased from the American Type Culture Collection (ATCC) include MDA-MB-231, NCI-H1299, A375, SKOV-3, OVCAR-3, A549, PC3, PANC-1, MDA-MB-468, A2780, HCC1937, and HCC70. Melanoma cell lines 63T, WM793, CSK-A375, and A2058 were sourced as previously described 28. Cell lines IGR-1, LOX-IMVI, and Malme-3M were kindly provided by Prof. Yardena Samuels (Weizmann Institute of Science, Rehovot, Israel). UM-SCC-47 line was a kind gift from Dr. Itay Tirosh (Weizmann Institute of Science, Rehovot, Israel). Melanoma cell line SB-2 was a kind gift from Prof. Menashe Bar-Eli (The University of Texas MD Anderson Cancer Center, Houston, TX, USA). Cancer cell lines MDA-MB-231, A375, CSK-A375, WM793, A2058, 63T, IGR-1 LOX-IMVI, SB-2, Malme-3M, SKOV-3, OVCAR-3, A549, PC3, PANC-1, MDA-MB-468 and A2780 were cultured in DMEM, supplemented with 10% FCS (Gibco) 2 mM glutamine, 2 mM NEAA, and 100 U/mL penicillin-streptomycin. Cell lines NCI-H1299, HCC1937, and HCC70 were cultured in an RPMI medium containing 10% FCS (Gibco), 2 mM glutamine, 2 mM NEAA and 100 U/mL penicillin-streptomycin. UM-SCC-47 cell line was cultured in three parts of Hams F-12 nutrient mixture medium and one part of DMEM with 10% FCS (Gibco) and 100 U/mL penicillin–streptomycin. All cell lines were maintained at 37°C under a humidified atmosphere of 5% CO2 and 95 % air. All cell lines were routinely tested for mycoplasma using the mycoplasma detection kit, MycoAlert™ (Lonza Nottingham, Ltd). Frozen vials of cell stock were thawed, and cells were cultured for no more than a month for the experiments. Additional information on all the cell lines used in this study is summarized in Table S1.
Transfection procedures
Knockdown experiments: Transfection was performed using siGENOME Human YAP1 siRNA, SMART pool; M-012200-00-0005 and siGENOME Human WWTR1 (TAZ) siRNA, SMART pool; M-016083-00-0005 (GE Healthcare Dharmacon). siRNA non-targeting pool #2 (GE Healthcare Dharmacon) was used as a control for all the knockdown experiments. The SMART pool and single oligo siRNA sequences used for the experiments are enlisted in the supplementary Table S2 and Table S3. The final concentration for each siRNA was 30 nmol/L. The siRNAs were transfected using DharmaFect transfection reagent, according to the manufacturer’s protocol. Fresh medium was added after 6 h to replace the medium with transfection reagent and oligonucleotides. Cells were incubated for 48 h following the siRNA transfection and proceeded for further experiments.
Overexpression experiments: IRES-GFP empty vector was sourced as previously described (Revach et al., 2019). pcDNA Flag YAP1 and pcDNA3 Flag TAZ were obtained as described previously 29. Flag YAP1 and Flag TAZ were transfected using the jetPEI transfection reagent (Polyplus-Transfection®) according to the manufacturer’s instructions. The plasmids were purified using Qiagen Maxi Kit and quantified using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Catalog No: ND-2000). Briefly, MDA-MB-231 and NCI-H1299 (0.7 × 106) cells were seeded on a 10 cm cell culture dish and incubated for 24 h. Subsequently, the cells were transfected with empty vector, Flag-YAP1, Flag-TAZ, and Flag-YAP1 plus Flag-TAZ. The final concentration of each plasmid used for overexpression was 10 µg per 10 cm culture dish. For Flag-YAP1 plus Flag-TAZ condition, the total concentration was 20 µg per 10 cm cell culture dish. The cells were then incubated for 48 h before proceeding with further experiments.
Preparation of gelatin-coated culture plates
Gelatin coating of culture plates for the matrix degradation assay was conducted as previously described 28. Briefly, ninety-six micro-well plates, with a glass-bottom (Thermo Fisher Scientific Catalog No-164588) were treated with 50 mg/mL of poly-lysine (Sigma Aldrich, Catalog No-P-7405) prepared as 1:1 mixture with Dulbecco’s Phosphate Buffered Saline (DPBS, Biological Industries, Catalog No-02-023-1A) and incubated for 20 minutes at room temperature. Then, the poly-lysine solution was removed, and the plate was washed (×3) with DPBS. Porcine skin gelatin (Sigma Aldrich, catalog no. G2500) was prepared in DPBS (2 mg/ml) and filtered through a 0.22-micron Steritop filter (Millipore Fisher Scientific Catalog No-15770319). Gelatin was fluorescently labeled using Alexa Fluor 488 Protein labeling kit (Molecular Probes, Thermo Fisher Scientific) according to the manufacturer’s instructions. The gelatin was subsequently cross-linked using N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC hydrochloride) (Sigma Aldrich Catalog No- 03450) and N-Hydroxysuccinimide (NHS) (Sigma Aldrich Catalog No-130672), prepared as 10% solutions in ddH2O. Subsequently, 96-well glass-bottom plates were coated with 40µl volume of a mixture (10:1) of unlabeled gelatin and the Alexa Fluor488 labeled gelatin. For every 100µl gelatin, the gelatin to cross-linker mixture ratios was (82.5µl gelatin: 12.5 µl NHS: 5µl EDC). Subsequently, the gelatin mixture was incubated for 1 h at room temperature. After incubation, the surfaces of the gelatin-coated plates were washed three times with DPBS. 100µl of fresh DPBS solution was added to the wells and UV sterilized for 30 minutes in the biosafety hood. The plates were then ready to use.
Gelatin degradation assay
The different cancer cell lines (104) were seeded on the Alexa Fluor 488-labeled gelatin matrix in the 96-well plates and cultured for 5-6 h. The cells were then fixed and stained for F-actin and DAPI, washed with DPBS, and kept wet for imaging. Every experiment was performed in two or more replicate wells for each condition. Z stack of images was acquired using a WiScan® Hermes Automated High Content Imaging System (IDEA Bio-Medical Ltd.) using 40X/0.75 NA air objective. In a single well, images from a total of 36 fields were acquired, and the cells were counted based on the nuclear (DAPI) staining. The degraded gelatin area (µm2 per cell) was calculated using the Image J software (rsbweb.nih.gov/ij). All degradation values following treatments were compared to those of control cells cultured on the same plate in every independent set of experiments.
The effect of EGF stimulation on gelatin degradation
Cancer cell lines (SKOV-3 and OVCAR-3) were grown in a medium containing 10 % FBS and serum-starved (0.5 % FBS) for 24 h. Subsequently, cells that were cultured in 0.5 % FBS were stimulated with hEGF (recombinant EGF) (Sigma Alrdich, Catalog No: E9644) in concentrations (10 and 30 ng/ml) and plated on Alexa 488-gelatin for 5-6 h. The cells were then processed for the gelatin degradation assay as described above.
The effect of YAP/TAZ inhibitor verteporfin on gelatin degradation
For gelatin degradation, an assay measuring the effect of the YAP/TAZ inhibitor verteporfin (Holland Moran, Catalog No 1711461), MDA-MB-231 and NCI-H1299 cells were seeded on the Alexa Fluor 488 labeled gelatin matrix and cultured for 2 h. Subsequently, the cells were treated either with vehicle (0.1% DMSO) or with different concentrations of verteporfin (final concentration range: 0.5-20 µM) for 4 h. The cells were then fixed and stained for TRITC-Phalloidin and DAPI and imaged using a WiScan® Hermes Automated High Content Imaging System (IDEA Bio-Medical Ltd.) under 10X/0.75 NA air objective.
Immunostaining, microscopy, and image analysis
Cells (104/well) were seeded on the Alexa Fluor488-tagged gelatin in 96-well glass-bottom plates. At the end of the gelatin degradation experiment (usually 5-6 hours after plating) the cells were fixed and permeabilized for 3 minutes with 3% PFA, 0.5% Triton-X-100, in DPBS followed by 3% PFA for another 30 minutes. Then, the cells were washed (×3) with DPBS and incubated with the primary antibody for 1 h. Subsequently, the cells were washed with DPBS (×3) and incubated with the appropriate secondary antibody for 30 minutes. The cells were washed again (×3) with DPBS and kept in DPBS for imaging. The images were acquired using a DeltaVision Elite microscopy system, equipped with a microtiter stage (Applied Precision Inc., Issaquah, WA) with 40X/0.75 air or 60X/1.42 oil objectives (Olympus). All acquired images were analyzed using Image J software (https://imagej.nih.gov/ij/).
Cell viability assay
A microscopy-based cell viability assay was performed as previously described 30. Hoechst 33342 (1 µg/ml; ImmunoChemistry Technologies, Bloomington, MN, USA) and propidium iodide (250 ng/ml; Sigma Aldrich, St. Louis, MO, USA) in DMEM were added onto cells and kept in the 37 °C incubator for 45 minutes. The cells were centrifuged at 1200 rpm for 3 minutes and then proceeded for imaging. The images were acquired uusing a WIScan Hermes® microscope with a 10X objective (IDEA Bio-Medical Ltd.), and the percentage of live and dead cells was calculated using WiSoft® Athena software (IDEA Bio-Medical Ltd.).
Western blot analysis
Cells plated on a 10 cm cell culture dish were scraped using a cell scraper and suspended in 300 µl ice-cold RIPA buffer. The cell lysates were kept on ice and vortexed at 5-minute intervals for over 45 minutes. Then, the lysates were cleared by centrifugation at 274 g at 4°C for 10 minutes. The cell lysates were either freshly examined or stored at –80°C. The frozen samples were thawed on ice, subjected to 10% polyacrylamide/SDS gel electrophoresis, and subsequently blotted onto Polyvinylidene fluoride (PVDF) membrane (Merck Millipore® Catalog No IPVH00010). The blots were blocked using 5% skimmed milk in Tris Buffered Saline, containing Tween 20, pH 8.0 (TBST) buffer, and probed with primary antibody overnight at 4 °C. They were then washed with TBST buffer X3 (10 minutes each) and incubated with HRP-coupled secondary antibody (Jackson ImmunoResearch Laboratories Inc.) for 1 h. Chemiluminescent Super Signal West Pico substrate (Thermo Fisher Scientific, Catalog number: 34579) was used for detecting the bands, and the blots were imaged using ChemiDoc MP Imager and quantified using Image Lab 4.1 software (BioRad, USA).
Quantitative real-time PCR
RNeasy Mini Kit (Catalog No-74104; Qiagen) was used for isolating total RNA from cells. Total RNA (1-2 µg) was reverse-transcribed using LunaScript™ RT SuperMix Kit (New England Biolabs, Catalog No; E3010S). Quantitative real-time RT-PCR (qRT-PCR) was performed using a Fast SYBR Green Master Mix using the OneStep instrument (Applied Biosystems). The obtained values were normalized to either HPRT1 or GAPDH genes. The primers used for the experiments are enlisted in Table S4.
Preparation of cells for proteomic profiling
For knockdown experiments, MDA-MB-231 (0.7× 106) cells were seeded on 10 cm cell culture dishes, incubated for 24 h, transfected with siControl, siYAP, siTAZ, and a mixture of siYAP and siTAZ SMARTpools, and further incubated for 48 h. The YAP/TAZ knockdown cells were seeded on unlabeled gelatin-coated 10 cm plates and cultured at 37ºC for 5-6 h. The effect of the treatment on the cells’ gelatin degradation activity was verified, in parallel, by plating a sample of the same cells on Alexa Fluor 488-gelatin plates and measuring their gelatin degradation phenotype. The experimental design consisted of three independent set of experiments for each knockdown condition. The cells were washed with 5 ml cold PBS and then scraped into 1 ml fresh ice-cold PBS, centrifuged at 3000 rpm at 4 °C, and the cell pellets were flash-frozen in liquid nitrogen and stored at -80 °C.
Sample preparation for proteomic analysis
Frozen cell pellet samples were dissolved in 5% SDS, 50 mM Tris-HCl (pH 7.5) The total protein concentration was measured using a BCA assay. 100µg of each sample was used for the downstream preparation. Dithiothreitol (DTT) was prepared fresh in 50 mM ammonium bicarbonate, and added to a final concentration of 5 mM. The samples were then incubated at 56°C for 1 hour. Iodoacetamide was prepared fresh in 50mM ammonium bicarbonate, and was added to final concentration of 10mM. Samples were incubated in the dark for 45 min. Phosphoric acid was then added to the samples to final concertation of 1%. The samples were mixed with 350 µL of 90% methanol along with 10% 50 mM ammonium bicarbonate, then transferred to the S-trap filter and centrifuged for 1 min at 4000xg and washed 3 times with 400µL of 90% MeOH+10% 50mM ammonium bicarbonate, then centrifuged at 4000g 1min. 4 µL of 0.5 µg/µL Trypsin in 125 µL in ammonium bicarbonate (25:1 protein amount: trypsin) was added to the samples. Samples were incubated at 37°C overnight. The next day, peptides were eluted using 80µL 50mM ammonium bicarbonate, which was added to the S-trap cartridge, centrifuged at 4000g for 1 min into new tubes, and collected the peptides. Then, a second digestion was performed using 4 µL of 0.5 µg/µL trypsin in 50mM ammonium bicarbonate was added to the eluted samples and incubated at 37°C for 4 hrs. Two more elations from the S-trap cartridge were performed. One with 80 µl of 0.2% formic acid, which was added to the S-trap cartridge and spun down at 4000g for 1 min. The second was done using 80 µL of 50% acetonitrile+0.2% formic acid was added to the cartridge and spun down at 4000g for 1 min. The three elutions were mixed and dried using a vacuum centrifuge (Centrivac, LabConco).
Proteomic analysis
The resulting peptides were analyzed using a nanoAcquity liquid chromatography (Waters) coupled with a Q Exactive HF-X (Thermo fisher scientific). Samples were analyzed randomly, loaded on a Symmetry C18 trap column (20mm X 0.18mm, 5um, Waters), and resolved on an HSS T3 (250mm X 0.075mm, 1.8um, Waters) analytical column at 350nl/min, using a gradient of 4-27%B (MeCN, 0.1% formic acid) for 155min. MS1 acquisition was performed at m/z range of 375-1650m/z at 120,000 resolution (@400m/z), allowing Automatic Gain Control (AGC) target of 106 with a maximum Injection Time (IT) of 60ms. MS2 acquisition was performed on the Top10 ions at Data-Dependent Acquisition (DDA) using Higher-energy Collisional Dissociation (HCD) fragmentation set at 27 Normalized Collision Energy (NCE) acquired at 15,000 resolution (@200 m/z). IT was set to 60ms and AGC to 1e5. Dynamic exclusion was set to 30sec with a counter of 1. The resulting data was processed with MaxQuant (v1.6.6.0). The data were searched with the Andromeda search engine against the Human proteome database (SwissProt Nov20) appended with common lab protein contaminants. The following modifications were allowed: fixed carbamidomethylation on C, variable protein N-terminal acetylation, variable deamidation on NQ and variable oxidation on M. The quantification was based on the LFQ method, based on unique peptides.
Bioinformatics analysis
For each cancer cell line used in this study (See Supplementary Table S1), we retrieved the information concerning the tissue of origin (e.g., primary tumor vs. metastases) from the CCLE database (https://depmap.org/portal/download/). Bioinformatic analysis of the proteomic data of MDA-MB-231 cells was applied on LFQ intensities of 4,704 detected proteins. Proteins having at least two one razor and unique peptides were considered, and 37 known contamination were removed from the analysis. For the detection of differential proteins, intensities were log2 transformed and analyzed with ANOVA following a multiple test correction (FDR step-up) using Partek Genomics Suite 7.0. For each pairwise comparison, we considered proteins having at least two valid measurements (out of 3) in both groups and that passed the thresholds of fold change
|(log2)| > 1 and p-value <0.16. In addition, proteins that were detected in at least 2 replicates in one group and completely absent in the other group were also considered as qualitatively differential proteins. For visualization of the protein expression, heat maps were prepared using Partek Genomics Suite, using log2-transformed LFQ intensities with row standardization (scaling the means of a row to zero, with a standard deviation of 1), and partition clustering using the k-means algorithm (Euclidian method). For volcano plot visualization, missing values were imputed to a value of 15, and new fold change values were calculated with ANOVA and visualized using MATLAB. Principle component analysis (PCA) was calculated using Partek Genomics Suite. For visualization of proteins network, the relations between differential proteins of the double knockdown was inferred with StringDB 31 as a “full STRING network”, and visualized by Cytoscape 3.7.2 32. The width of the edges corresponds to the “combined score” or StringDB, and the protein color scale corresponds to their log2 fold change as inferred from the proteomics ANOVA. Proteins that change qualitatively (were detected only in one condition) were assigned an imputed value of +/- 5. The assignment of proteins belonging to Hippo signaling, cell adhesion, and ECM remodeling pathways was inferred using the GeneCards suite 33. A list of invadopodia-related proteins was compiled by data mining in the Harmonizome database and the related literature 34, 35, 36, 37, 38, 39, 40, 41, 42, 43.
RNA sequencing
MDA-MB-231 cells (0.7 ×106) were seeded on 10 cm culture dishes and incubated for 24 h. Subsequently, the cells were transfected with SMARTpool siRNA for siControl, siYAP, siTAZ, and siYAPTAZ and incubated for a further 48 h for the knockdown. After the incubation, the knockdown cells were seeded on non-labeled gelatin-coated 10 cm plates and cultured for 5-6 h in an incubator. Subsequently, the culture medium was aspirated, and the cells were given a wash with DPBS. One ml of DPBS was added to the plate, and the cells were then scraped using a cell scraper. Then the cells were centrifuged at 3000 rpm at 4 ° C. The supernatant was aspirated, and the RNA was extracted from the cell pellet. The RNA extraction was done using RNeasy Mini Kit (Catalog Nos; 74104, 74106 Qiagen) according to the manufacturer’s instructions. The RNA was quantified using Qubit 3 Fluorometer (ThermoFisher scientific, USA) and TapeStation (Agilent Technologies 4200, USA) to assess the purity of RNA. A RIN (RNA Integrity Number) score ranging from 9-10 was obtained for each condition. Then RNA-seq libraries were prepared at the Crown Genomics institute of the Nancy and Stephen Grand Israel National Center for Personalized Medicine, Weizmann Institute of Science. Libraries were prepared using the INCPM-mRNA-seq protocol. Briefly, the polyA fraction (mRNA) was purified from 500 ng of total input RNA followed by fragmentation and the generation of double-stranded cDNA. Afterwards, Agencourt Ampure XP beads cleanup (Beckman Coulter), end repair, A base addition, adapter ligation and PCR amplification steps were performed. Libraries were quantified by Qubit (Thermo fisher scientific) and TapeStation (Agilent). Sequencing was done on a NovaSeq6000 instrument (Illumina) using an SP 100 cycles kit (single read sequencing).
Transcriptomic analysis
RNA sequencing analysis was done using the UTAP transcriptome analysis user-friendly Transcriptome Analysis Pipeline (UTAP v1.10) transcriptome analysis pipeline (Kohen, Barlev et al. 2019). Reads were trimmed to remove adapters and low quality bases using cut adapt (-a “A pipeline 44. Reads were trimmed to remove adapters and low quality bases using cut adapt (-a “A (10)” -a “T (10)” –times 2 -q 20 -m 25) 45 and mapped to the human genome (GRCh38, GENECODE version 34) using STAR v2.4.2a 46 (using–alignEndsType EndToEnd, –outFilterMismatchNoverLmax. 0.05, –two pass Mode Basic). Reads were counted using STAR, and genes having minimum of five reads in at least one sample were considered. Normalized counts and detection of differential expression were performed using DESeq2 47 (betaPrior, cooksCutoff, and independent filtering parameters set to False). Differentially expressed genes were selected with absolute fold change (log2)≥1 and adjusted multiple testing p-value ≤0.05 48. Matlab was used to generate the volcano plots.
Data Availability
The raw data of proteomic profiling has been deposited at the ProteomeXchange via the Proteomic Identification Database (PRIDE partner repository). The dataset identifier ID is PXD034562. The accession numbers for the RNA-seq data have been deposited in NCBI's Gene Expression Omnibus (Edgar et al., 2002) and are accessible through Genome Sequence Archive for Human under GEO Series accession number: GSE205726. There are no restrictions on the availability of the data.
Statistical analysis
All statistical analysis of the experimental data was performed using the Graph Pad Prism version 8.0.1 for Windows, GraphPad Software, San Diego, California USA, www.graphpad.com software. Statistical significance for each experiment is marked in the form of asterisks (*) along with the calculated p-value for experiments are shown.